- Email: [email protected]
Incidence and progression of cataract in the Melbourne Visual Impairment Project
Incidence and Progression of Cataract in the Melbourne Visual Impairment Project CATHERINE ANNE MCCARTY, PHD, MPH, BICKOL NANJAN MUKESH, PHD, PETER N. DIMITROV, BORTHOP, AND HUGH RINGLAND TAYLOR, MD, FRACO
● PURPOSE: To determine the 5-year incidence and progression of cataract and cataract surgery in the Melbourne Visual Impairment Project. ● DESIGN: Prospective cohort study. ● METHODS: Demographic information including race, sex, age, and education level was collected at baseline. Cortical cataract was defined as 4/16 or greater opacity; progression was defined as a more than 2/16 increase. Nuclear cataract was defined as Wilmer standard grade 2 or higher; progression was defined as more than 0.5 increase. Posterior subcapsular (PSC) cataract was defined as opacity 1 mm2 or greater; progression was defined as greater than 1 mm2 increase. ● RESULTS: Of the 3,040 participants eligible to attend follow-up examinations, 2,594 (85% of those eligible) participated. The mean age of participants at follow-up was 62.5 years, and 55% were female. The percentage of patients who had at least one lens extracted over 5 years increased from 0.5% of those aged 40 to 49 years at baseline to 35.7% of those aged 80 years or more at baseline. The overall incidence of the three types of cataract was as follows: cortical 7.7% (95% confidence limits [CL] ⴝ 5.8 –9.8), nuclear 16.4% (95% CL ⴝ 12.1–20.8), and PSC 7% (95% CL ⴝ 5.3– 8.7). The overall progression of cataract was cortical 14.3% (95% CL ⴝ 10.2–18.3), nuclear 19.3% (95% CL ⴝ 15.9 – 22.7), and PSC 20% (95% CL ⴝ 8.7–31.1). The
Accepted for publication Aug 23, 2002. From the Marshfield Medical Research Foundation (C.A.M.), Marshfield Clinic, Marshfield, Wisconsin; and Centre for Eye Research Australia (C.A.M., B.N.M., P.N.D., H.R.T.), University of Melbourne, Melbourne, Australia. This study was supported by the National Health and Medical Research Council, the Victorian Health Promotion Foundation, the Dorothy Edols Estate, the Ansell Ophthalmology Foundation, the Jack Brockhoff Foundation, the Ian Potter Foundation, the Eye Ear Nose and Throat Research Institute, the Felton Bequest, the Hugh D. Williamson Foundation, and the Appel Family Bequest. Doctor McCarty was the recipient of the Wagstaff Research Fellowship in Ophthalmology from the Royal Victorian Eye and Ear Hospital. Inquiries to Catherine McCarty, PhD, MPH, Marshfield Medical Research Foundation, Marshfield Clinic, 1000 N. Oak Ave (ML2), Marshfield, WI 54449; fax: (715) 389-3880; e-mail: mccartyc@mmrf. mfldclin.edu
10
©
2003 BY
incidence and progression rates increased significantly by age, but the rates were not significantly different by sex. ● CONCLUSIONS: These cataract incidence data confirm the public health importance of cataract in Australia. The data also support the need to plan both primary prevention program and adequate surgical services to meet the anticipated increase in demand with the aging population. (Am J Ophthalmol 2003; 136:10 –17. © 2003 by Elsevier Inc. All rights reserved.)
C
ATARACT SURGERY IS THE MOST COMMONLY PER-
formed ophthalmic procedure in Australia,1 and the amount of cataract surgery is likely to double in the next 20 years with the aging of the population. We have shown previously in the Visual Impairment Project, a cross-sectional, population-based study, that although there are a number of identifiable risk factors for the prevalence of cataract,2 no segment of the adult Melbourne population was observed to be systematically underserviced in terms of cataract surgery.3 From these same data we estimated that the projected need for cataract surgery varies by level of opacity, visual acuity, and patient concern about vision.4 Relatively few data are available regarding the incidence of cataract5–10 or cataract surgery.11 The incidence of cataract has been shown in these four studies to be highly age dependent, but the studies differ as to whether incidence rates are higher among male or female populations. Data from the Beaver Dam Eye Study indicate that the location of lens opacity in the posterior subcapsular (PSC) region, as opposed to the cortex or nucleus, is the most important predictor of subsequent cataract extraction.11 As acknowledged by all of the authors, incidence data are useful for planning future health service delivery needs, in conjunction with population projections, and also for planning future intervention studies. The purpose of this study was to quantify the 5-year incidence of cataract and cataract surgery in the Melbourne Visual Impairment Project cohort.
ELSEVIER INC. ALL
RIGHTS RESERVED.
0002-9394/03/$30.00 PII S0002-9394(02)01844-5
METHODS THE MELBOURNE VISUAL IMPAIRMENT PROJECT IS A POPU-
lation-based study of eye disease, the methods for which have been published previously.12 Recruitment and baseline examinations for this prospective cohort study were conducted from 1992 to 1994. Nine pairs of census collector districts from the Melbourne Statistical Division were randomly selected from which to recruit eligible residents via a household census. Eligible residents were defined as permanent residents aged 40 years and older. At the time of the household census, basic demographic details including age, sex, race, education level, country of birth, language spoken at home, and use of eye care services, were collected. Appointments were scheduled at locally established test sites. The protocol was approved by the Human Research and Ethics Committee of the Royal Victorian Eye and Ear Hospital and conforms to the Declaration of Helsinki for research involving human subjects. In 1997, a 5-year follow-up examination commenced.13 All participants were traced, and participants living in the area were invited to again attend locally established test sites for the follow-up examinations. The same examination protocol was used with redundant questions eliminated. Reported deaths were confirmed through the National Death Index at the Australian Institute for Health and Welfare in Canberra. The following information was collected for each participant: presenting and best-corrected visual acuity on a 4 m logarithm of the minimal angle of resolution (LogMAR) chart, LogMAR reading vision at a distance comfortable for the participant, Humphrey Fastpac 24-2 visual fields (Humphrey Instruments Inc., San Leandro, California, USA), Tonopen (Oculab, Calendale, California, USA) intraocular pressure, personal health history, clinical slit-lamp examination, and photography of the lens and fundus. The clinical examination and photography of the lens were performed after pupil dilation to a minimum of 6 mm with tropicamide. Cataract was graded clinically and on photographs according to the Wilmer protocol.14 The cortex was assessed on retroillumination and the nucleus was assessed with a slit lamp. Cortical cataract was defined as 4/16 or greater opacity. Nuclear cataract was defined as Wilmer standard grade 2.0 or higher. Posterior subcapsular cataract was defined as PSC opacity greater than or equal to 1 mm2. Photogrades were used when available, and the agreement between photogrades and clinical grades was shown to be very high for all three types of opacity.2 Participants were queried about the year that any cataract extraction had been performed. The treating ophthalmologist was contacted to provide information on the type of opacity present before cataract extraction. Prior cataract surgery was confirmed on clinical examination and the status of the capsule, if present, was noted. VOL. 136, NO. 1
CATARACT INCIDENCE
IN
Patients were classified as having incidence cataract or progression of lens opacity on the basis of that event occurring in either eye, regardless of the status of the fellow eye at baseline. Except where indicated in the results, the specific cataract types are not mutually exclusive, that is, a person may be classified as having any, none, or all three of the cataract types. For each of the separate cataract types, the denominator for the incidence estimates included those participants who did not have that specific cataract type (as defined in the previous paragraph) at baseline in either eye. Given these definitions, it is possible that someone was classified as having incident cataract, but not classified as having opacity that progressed because the progression was not great enough to fit into the definition for progression. For each specific type of cataract, the numerator for incidence estimates included participants who developed a specific cataract type over the time period as well as those who had cataract surgery (reportedly for that type of cataract according to the treating ophthalmologist) during the time period. Progression of lens opacity was considered separately to cataract. Thus, a patient did not necessarily have to have a “cataract” to be classified as having a lens opacity that had progressed. Patients were considered to be at risk of having their lens opacity progress if they had opacity greater than 0 at baseline, if they did not have cataract surgery in both eyes over the intervening time period and if they did not have the highest level of opacity at baseline so that there was no possibility of progression. Cataract surgery was not included in the definition of progression of lens opacity. To assess measurement error associated with photograding of the lens, a sample of photographs from baseline was regraded by one of the photograders (P.N.D.) at follow-up. Interview data were entered directly into a Paradox database, while clinical data were entered twice and verified. SAS (SAS Software, Cary, North Carolina, USA) was used for all statistical analyses, and P ⬍ .05 was considered statistically significant. Intraclass correlation coefficients were calculated to assess agreement between graders. Ninety-five percent confidence limits (CL) around the incidence estimates were calculated according to Cochran15 to account for the cluster sampling design.
RESULTS TWO-HUNDRED THIRTY-ONE (7%) OF THE PARTICIPANTS
examined at baseline had died before the 5-year follow-up examinations.14 Participants with presenting visual acuity less than 6/12 at baseline were 2.3 times as likely to have died in the intervening 5 years. Eighty-five percent (2,594/ 3,040) of the remaining eligible group participated in the incidence study. The mean age of the participants at follow-up was 62.5 years (SD ⫽ 10.9 years) and 1,421 VISUAL IMPAIRMENT STUDY
11
12
AMERICAN JOURNAL
OF
No cataract-no cataract No cataract-cataract No cataract-surgery Cataract-cataract Cataract-surgery Surgery-surgery Total
*Participants potentially at risk for incident cataract. Normal font ⫽ cortical, bold font ⫽ nuclear, italic font ⫽ posterior subcapsular. The shaded area indicates participants potentially at risk for incident cataract.
2,786, 2,871, 3,033 158, 68, 59 32, 27, 47 179, 184, 31 25, 30, 10 51, 51, 51 3,231, 3,231, 3,231 565, 593, 620 45, 17, 13 7, 5, 9 35, 28, 6 5, 7, 3 0, 0, 0 657, 650, 651 148, 158, 196 26, 8, 6 6, 6, 11 30, 38, 2 7, 7, 2 12, 12, 12 229, 229, 229 23, 38, 43* 11, 4, 7* 10, 5, 14* 24, 16, 8 7, 12, 3 39, 39, 39 114, 114, 114 1, 9, 9* 4, 2, 3* 1, 4, 3* 11, 16, 4 6, 4, 2 0, 0, 0 23, 35, 21 67, 227, 51* 26, 29, 5* 0, 0, 0* 79, 86, 11 0, 0, 0 0, 0, 0 172, 342, 67 63, 77, 83* 42, 6, 22* 0, 0, 0* 0, 0, 0 0, 0, 0 0, 0, 0 105, 83, 105 1,897, 1,756, 2,009* 0, 0, 0* 0, 0, 0* 0, 0, 0 0, 0, 0 0, 0, 0 1,897, 1,756, 2,009
Baseline
22, 13, 22* 4, 2, 3* 8, 7, 10* 0, 0, 0 0, 0, 0 0, 0, 0 34, 22, 35
SurgerySurgery CataractSurgery CataractCataract No CataractSurgery No CataractCataract
Follow-up
OPHTHALMOLOGY
No CataractNo Cataract
TABLE 1. Lens Status in Each Eye of Participants at Baseline and Follow-up
Died
Missing Data for Both Eyes
Total
(55%) were female. The mean time elapsed between baseline and follow-up examinations was 4.5 years (range 4 –7; SD ⫽ 0.64). Of the 2,594 participants examined at follow-up, 2,392 (92%) had lens data available for both eyes at follow-up. The status of lens data for the three types of opacities in all 3,231 participants with baseline lens data are summarized in Table 1. The shaded area indicates those who were at risk of incident cataract. However, the actual number of persons at risk for a particular type of cataract is slightly less owing to data missing for the only eye at risk of incident cataract. For example, if a person had no cortical cataract in one eye and previous cataract surgery in the other eye at baseline, and the first eye did not have a gradeable photograph at follow-up, this person would not be included in the denominator for incident cortical cataract at follow-up. It is not possible from the information presented in Table 1 to infer the number of people at risk for progression of lens opacity, because some participants with no cataract at baseline had some lens opacity and were at risk of progression. The intraclass correlation for the agreement between the graders of baseline and follow-up cortical opacities was 0.95, and the agreement between the two graders of the follow-up cortical opacities was 0.85. The intraclass correlation for the agreement of nuclear opacity grading between baseline and follow-up was 0.78, whereas the agreement between the two graders at follow-up was 0.93. The intraclass correlation for the agreement between baseline and follow-up grading of PSC opacities was 0.86. The progression of the three types of lens opacities was defined to be greater than the largest (that is, most conservative) standard error for either of the graders, which was greater than the standard error of the difference between the graders: a change of more than 2/16 for cortical, a change of more than 0.5 for nuclear, and a change of more than 1 mm2 for PSC. The overall rate of cataract extraction in at least one eye was 5.1% (95% CL ⫽ 3.9 – 6.2). The rate increased significantly by age (Figure 1), but not by sex (P ⬎ .05, data not presented). The distribution of incident cataract type varied by age somewhat; however, pure incident nuclear cataract was the most common type in all age categories, followed by mixed cataract (Figure 1). The rate of cataract extraction was fairly steady since the baseline examination (data not presented), suggesting that participation in the study did not have any significant impact on cataract extraction. The rate of cataract extraction increased nearly linearly with the level of opacity at baseline and was higher overall for PSC opacity than cortical or nuclear opacity (Figures 2– 4). Of the 2,294 participants without cataract in either eye at baseline who had data at follow-up, 614 (26.8%) had developed cataract at follow-up. As we found at baseline,2 there was relatively little overlap of cataract type given the definition of cataract that was employed. Pure nuclear
JULY 2003
FIGURE 1. Age-specific incidence of any cataract or cataract surgery in at least one eye. PSC ⴝ posterior subcapsular.
FIGURE 2. Cumulative 5-year incident cataract extraction by level of cortical opacity at baseline.
FIGURE 3. Cumulative 5-year incident cataract extraction by level of nuclear opacity at baseline.
cataract was the predominant type, followed by pure cortical cataract (Figure 5). The age- and sex-specific incidence and progression rates of mixed cortical (Table 2), nuclear (Table 3), and PSC cataract (Table 4) reveal significant increases by age. VOL. 136, NO. 1
CATARACT INCIDENCE
IN
FIGURE 4. Cumulative 5-year incident cataract extraction by level of posterior subcapsular (PSC) opacity at baseline.
FIGURE 5. The distribution and overlap of the three cataract types in the 614 incident cases of cataract. PSC ⴝ posterior subcapsular.
Although some of the incidence and progression rates were higher among women, these differences were not statistically significant. Overall, the incidence of nuclear cataract was twice that of cortical or PSC cataract and this difference was statistically significant. The overall progression rates of the three types of cataract were not significantly different. The progression rate was higher than the incidence rate. There were only 45 cases of PSC cataract at baseline with available follow-up data—too few to calculate age- and sex-specific rates. The overall progression rate in this group was 20.0% (95% CL ⫽ 8.7–31.3). We calculated the incidence of cortical and nuclear cataract using two alternative definitions for each cataract type. The incidence of cortical cataract 2/16 or greater was 18.3% (95% CL ⫽ 14.1–22.5), 3/16 or greater was 15.8% (95% CL ⫽ 12.9 –18.6), and 6/16 or greater was 5.3% (95% CL ⫽ 3.9 – 6.7). The incidence of nuclear cataract standard 1.5 or greater was 45.3% (95% CL ⫽ 41.4 – 49.2) and standard 2.5 or greater was 8.9% (95% CL ⫽ 5.9 – 12.0). We were unable to conduct sensitivity analyses for PSC owing to the relatively small number of cases. VISUAL IMPAIRMENT STUDY
13
TABLE 2. Age and Sex-specific Incidence and Progression of Cortical Cataract Men Age Group (years)
Incidence 40–49 50–59 60–69 70–79 80⫹ Total Progression 40–49 50–59 60–69 70–79 80⫹ Total
Number at Risk
Women Percent (95% CL)
259 322 265 109 15 970
1.9 (0.34–3.5) 4.0 (2.3–5.7) 8.3 (4.5–12.2) 11.9 (3.9–19.9) 13.3 (0–29.9) 5.7 (3.6–7.8)
8 28 52 55 9 152
12.5 (0.57–24.4) 14.3 (0–29.9) 9.6 (0.03–19.2) 12.7 (5.4–20.1) 11.1 (0–33.6) 11.8 (5.7–18.0)
All
Number at Risk
Percent (95% CL)
Number at Risk
Percent (95% CL)
343 401 278 100 21 1,143
3.5 (0.84–6.2) 9.0 (6.3–11.6) 16.2 (13.2–19.2) 12.0 (5.1–18.9) 9.5 (0–22.2) 9.4 (7.5–11.2)
602 723 543 209 36 2,313
2.8 (1.1–4.5) 6.8 (4.7–8.8) 12.3 (10.5–14.2) 12.0 (6.7–17.3) 11.1 (2.5–19.7) 7.7 (5.8–9.5)
12 38 89 54 27 220
8.3 (0–24.0) 23.7 (13.9–33.5) 14.6 (7.3–21.9) 18.5 (5.4–31.7) 7.4 (0–16.3) 15.9 (11.0–20.8)
20 66 141 109 36 372
10.0 (0.88–19.1) 19.7 (11.6–27.8) 12.8 (6.5–10.1) 15.6 (8.4–22.8) 8.3 (0.18–16.5) 14.3 (10.2–18.3)
CL ⫽ confidence limits.
TABLE 3. Age and Sex-specific Incidence and Progression of Nuclear Cataract
Age Group (years)
Incidence 40–49 50–59 60–69 70–79 80⫹ Total Progression 40–49 50–59 60–69 70–79 80⫹ Total
Men Number at Risk
Women
All
Percent (95% CL)
Number at Risk
Percent (95% CL)
Number at Risk
Percent (95% CL)
259 330 273 114 9 985
1.5 (0.44–2.7) 7.0 (3.1–10.8) 26.0 (18.7–33.3) 55.3 (47.9–62.7) 66.7 (36.4–96.9) 17.0 (13.0–20.9)
343 405 284 89 13 1,134
2.0 (0.13–4.0) 7.7 (3.0–12.3) 28.9 (22.6–35.2) 57.3 (46.2–68.4) 76.9 (49.3–10.5) 15.9 (10.6–21.4)
602 735 557 203 22 2,119
1.8 (0.48–3.2) 7.4 (3.4–11.3) 27.5 (21.6–33.4) 56.2 (50.3–62.1) 72.7 (51.8–93.6) 16.4 (12.1–20.8)
259 330 283 134 18 1,024
7.7 (4.4–11.1) 12.7 (8.8–16.6) 20.9 (16.6–25.1) 47.0 (37.7–56.3) 50.0 (20.0–80.0) 18.9 (16.0–21.7)
345 405 303 123 33 1,209
10.4 (7.4–13.4) 12.6 (8.6–16.6) 26.1 (19.5–32.6) 47.9 (39.3–56.7) 39.4 (23.5–55.3) 19.7 (15.3–24.1)
604 735 586 257 51 2,233
9.3 (6.6–12.0) 12.7 (9.1–16.3) 23.6 (18.7–28.4) 47.5 (39.9–55.0) 43.1 (35.4–50.8) 19.3 (15.9–22.7)
CL ⫽ confidence limits.
levels of ultraviolet (UV-B) radiation in these environments. We have shown previously that higher levels of UV-B are significantly related to the prevalence of cortical cataract.2 Unexplainably, all of the observed cortical cataract incidence rates are substantially less than the 28.2% incidence rate that was observed among older participants aged 65 to 74 in the Italian-American Cataract Study.5 The cumulative nuclear cataract standard 2.0 or greater incidence rate of 16.4% observed in the current study is higher than in previous studies (Table 5). These differences may be explained by the different definitions of
DISCUSSION THESE DATA CONFIRM THAT CATARACT IS A MAJOR PUB-
lic health problem in Australia that will only increase as the population ages. The cumulative incidence rate of cortical cataract of 7.7% when defined as 4/16 or greater that was observed in the current study is similar to that of the Beaver Dam Eye Study,8 and the Longitudinal Study of Cataract7 (Table 5). When considering the cutoff of 3/16 or more, the incidence rate in the Visual Impairment Project is closer to that of the white participants in the Barbados Eye Study.10 That could be due to the higher 14
AMERICAN JOURNAL
OF
OPHTHALMOLOGY
JULY 2003
TABLE 4. Age and Sex-specific Incidence of Posterior Subcapsular Cataract
Age Group (years)
Men Number at Risk
40–49 50–59 60–69 70–79 80⫹ Total
258 329 274 127 16 1,004
Women Percent (95% CL)
All
Number at Risk
Percent (95% CL)
Number at Risk
Percent (95% CL)
344 404 297 118 31 1,194
1.2 (0.41–1.9) 4.2 (2.4–6.0) 11.1 (7.5–14.7) 21.2 (15.7–26.7) 16.1 (7.0–25.3) 7.0 (5.7–8.4)
602 733 571 245 47 2,198
2.3 (1.4–3.3) 4.1 (2.4–5.8) 8.1 (5.7–10.4) 22.9 (16.5–29.2) 17.0 (7.5–26.5) 7.0 (5.3–8.7)
3.9 (1.8–6.0) 4.0 (0.39–7.5) 4.7 (2.4–7.1) 24.4 (14.5–34.3) 18.8 (0.34–37.2) 7.0 (4.4–9.5)
CL ⫽ confidence limits.
TABLE 5. Published Incidence Rates of Cataract Study
Study Population
Time Frame
Cataract Definition
Melbourne Visual Impairment Project
Cluster, random sample of Melbourne residents aged 40⫹ at baseline, 1992– 1994
5 years
Italian-American Cataract Study Group5 Longitudinal Study of Cataract6,7
Follow-up of clinic-based case-control study in Parma, Italy, patients aged 45–79 at baseline, 1987–1989 Follow-up of clinic-based Lens Opacities Case-Control Study, patients aged 40–70⫹ at baseline, 1989–1993 Rural Wisconsin residents aged 43–84 at baseline, 1988–1990
5 years
Cortical ⱖ2/16 Cortical ⱖ3/16 Cortical ⱖ4/16 Cortical ⱖ6/16 Nuclear ⱖstandard 1.5 Nuclear ⱖstandard 2.0 Nuclear ⱖstandard 2.5 PSC ⱖ1 mm2 Any cataract Any and VA ⱕ02 LogMAR Cortical, any, aged 65–74 years Nuclear, any, aged 65–74 years PSC, any, aged 65–74 years Cortical, LOCS III ⱖ2.0 Nuclear PSC Cortical, ⱖ5% lens surface Nuclear, ⱖgrades 4 and 5 PSC, ⱖ5% grid segment Any lens opacity and VA ⱕ0.2 LogMAR
18.3% (14.1–22.5) 15.8% (12.9–18.6) 7.7% (5.8–9.5) 5.3% (3.9–6.7) 45.3% (41.4–49.2) 16.4% (12.1–20.8) 8.9% (5.9–12.0) 7.0% (5.3–8.7) 28.1% (23.5–32.7) 10.6% (8.3–13.0) 28.2% (SE ⫽ 3.2) 11.5% (SE ⫽ 1.5) 9.6% (SE ⫽ 1.3) 7.7% (SE ⫽ 1.2) 7.7% (SE ⫽ 1.5) 4.3% (SE ⫽ 0.9) 8.0% (not available) 13.1% (not available) 3.4% (not available) 9.0% (6.7–11.3)
Cortical, LOCS II scores ⱖ2 Nuclear, LOCS II scores ⱖ2 PSC, LOCS II scores ⱖ2
Black:22.2, White:16.2 Black:9.2, White:8.7 Black:3.3, White:3.9
Beaver Dam Eye Study8 Italian Study9
Barbados Eye Study10
5 years
5 years
Priverno participants in cardiovascular study, aged 45–69 years at baseline, 1987 Population-based sample of Barbados residents aged 40–84 at baseline, 1987–1992.
7 years
4 years
Cataract Rate (95% CL)
LOCS ⫽ Lens Opacity Case-Control Study; LogMAR ⫽ logarithm of the minimum angle of resolution; PSC ⫽ posterior subcapsular; VA ⫽ visual acuity.
nuclear cataract in the various studies, because when a more conservative definition was employed with our data, the incidence rate in the current study is similar to that in other studies. The cumulative PSC incidence rate of 7.0% in the current study is slightly higher than has been observed in VOL. 136, NO. 1
CATARACT INCIDENCE
IN
previous studies (Table 5). This difference may again reflect grading and classification differences. The data in Table 5, including the sensitivity analyses for the current study, reveal the large effect on incidence of different cut-off points for cataract classification. These data highlight the need for an objective cataract grading scheme and VISUAL IMPAIRMENT STUDY
15
definition that can be employed by all researchers to allow more valid comparisons between studies. The 28% 5-year incidence of any cataract in our population is significantly higher than the cumulative 7-year incidence of 9% reported in a sample of Italians aged 45 to 69 years.9 However, the definition of cataract employed by these researchers included a criterion of decreased visual acuity associated with age-related lens opacity. When we included the additional criterion of visual acuity less than 0.2 LogMAR in our definition of cataract, we found the incidence of any cataract to be 10.6% (95% CL ⫽ 8.3–13.0), which is not significantly different from the finding in the Italian study.9 Women have been reported to have significantly higher incidence rates of nuclear cataract in the Beaver Dam Eye Study8 and the Barbados Eye Study,10 whereas the Longitudinal Study of Cataract6 failed to find any significant difference in incidence of nuclear cataract between sexes. Although the incidence of nuclear cataract was slightly higher among women than men in our study, this finding was not statistically significant. Prior research has suggested that the increased prevalence of nuclear cataract that has been observed among women may be due to hormonal differences and that hormone replacement therapy may be protective against nuclear cataract;17 however, our baseline data did not support this finding2 and no prospective data are available. The threefold difference in cataract incidence that was observed in the Visual Impairment Project with varying definitions of cataract, as well as the difference in rates among studies with similar populations, highlights the lack of objectivity in current definitions of cataract. This problem is further compounded by the use of many different cataract grading schemes. One solution to this problem could be the inclusion of a visual acuity or visual function criterion along with lens opacity to define cataract. Another possible solution would be the use of objective systems, such as digital measurements, to classify lens opacity. Clearly, standardization of measurement technique and objectivity in defining cataract are necessary to allow comparisons of results between studies. The strengths of this study include the sampling strategy and relatively high response rate, which resulted in a study sample that is representative of the Melbourne population aged 40 years and older. A weakness of any incidence study with no data at intervening time points is that the estimation of incidence and progression could be affected by differential mortality. That would lead to a slight underestimate of the incidence and progression rates if participants who died were more likely to develop cataract or to have their opacity progress before their death. We showed previously that participants with cortical cataract at baseline were slightly more likely to die in the intervening 5 years, but this finding was not significant in multivariate analyses.14 An advantage of incidence data for cataract surgery over cross-sectional data is that the 16
AMERICAN JOURNAL
information can better be used to plan for eye care services because it is easier to estimate annual demand. Another limitation of these data is that information about the status of lenses immediately before cataract extraction was not available. Estimates of cataract surgery rates could also be influenced by participation in the baseline examination. Specifically, patients may be more likely to seek cataract extraction if they had been told during the ophthalmic examination that they had significant lens opacity. However, we found no obvious impact of participation in baseline examinations because the cataract extraction rates were fairly constant over the time between baseline and follow-up examinations. That is not unexpected, as we have shown previously in this cohort that no one under the age of 70 years had vision less than 6/12 due to cataract, and the rate of vision impairment due to cataract rose to 11% of patients aged 90 years and older.16 In conclusion, these cataract incidence data confirm the public health importance of cataract in Australia. The data also support the need to plan both primary prevention programs to encourage reduction in known environmental risk factors such as cigarette smoking and ocular UV-B radiation exposure and adequate surgical services to meet the anticipated increase in demand with the ageing population. ACKNOWLEDGMENTS
The authors acknowledge the assistance of Sharon Bayley, Marie Bissinella, Shayne Brown, Dr. Van Lansingh, Dr. Patricia Livingston, Claire McKean, Dr. Yury Stanislavsky, Dr. Mylan Van Newkirk, Cathy Walker, and Matthew Wensor in the data collection.
REFERENCES 1. Keeffe JE, Taylor HR. Cataract surgery in Australia 1985– 1994. Aust N Z J Ophthalmol 1996;24:313–317. 2. McCarty CA, Mukesh BN, Fu CL, Taylor HR. The epidemiology of cataract in Australia. Am J Ophthalmol 1999; 128:446 –465. 3. McCarty CA, Keeffe JE, Taylor HR. The need for cataract surgery: projections based on lens opacity, visual acuity, and personal concern. Br J Ophthalmol 1999;83:62–65. 4. McCarty CA, Nanjan MB, Taylor HR. Operated and unoperated cataract in Australia. Clin Exp Ophthalmol 2000;28: 77–82. 5. Italian-American Cataract Study Group. Incidence and progression of cortical, nuclear, and posterior subcapsular cataracts. Am J Ophthalmol 1994;118:623–631. 6. Leske MC, Chylack LT, Wu S-Y, et al. Incidence and progression of nuclear opacities in the Longitudinal Study of Cataract. Ophthalmology 1996;103:705–712. 7. Leske MC, Chylack LT, He Q, et al. Incidence and progression of cortical and posterior subcapsular opacities. The OF
OPHTHALMOLOGY
JULY 2003
8.
9.
10.
11.
12.
Longitudinal Study of Cataract. Ophthalmology 1997;104: 1987–1993. Klein BEK, Klein R, Lee KE. Incidence of age-related cataract. The Beaver Dam Eye Study. Arch Ophthalmol 1998;116:219 –225. Cedrone C, Culasso F, Cesareo M, et al. Prevalence and incidence of age-related cataract in a population sample from Priverno, Italy. Ophthalmic Epidemiol 1999;6:95–103. Leske MC, Wu S-Y, Nemesure B, Li X, Hennis A, Connell AMS, and the Barbados Eye Studies Group. Incidence and progression of lens opacities in the Barbados Eye Studies. Ophthalmology 2000;107:1267–1273. Klein BEK, Klein R, Moss SE. Incident cataract surgery. The Beaver Dam Eye Study. Ophthalmology 1997;104:573– 580. Livingston PM, Carson CA, Stanislavsky YL, Lee SE, Guest
VOL. 136, NO. 1
CATARACT INCIDENCE
IN
13.
14. 15. 16.
17.
CS, Taylor HR. Methods for a population-based study of eye disease: the Melbourne Visual Impairment Project. Ophthalmic Epidemiol 1994;1:139 –148. McCarty CA, Nanjan MB, Taylor HR. Vision impairment predicts five year mortality. Br J Ophthalmol 2001;85:322– 326. Taylor HR, West SK. A simple system for the clinical grading of lens opacities. Lens Res 1988;5:139 –148. Cochran WG. Sampling techniques. New York: John Wiley & Sons, 1977, 249 –271. Weih LM, VanNewkirk MR, McCarty CA, Taylor HR. Age-specific causes of bilateral visual impairment. Arch Ophthalmol 2000;118:264 –269. Klein BEK, Klein R, Ritter L. Is there evidence of an estrogen effect on age-related lens opacities? The Beaver Dam Eye Study. Arch Ophthalmol 1994;112:85–91.
VISUAL IMPAIRMENT STUDY
17
● PURPOSE: To determine the 5-year incidence and progression of cataract and cataract surgery in the Melbourne Visual Impairment Project. ● DESIGN: Prospective cohort study. ● METHODS: Demographic information including race, sex, age, and education level was collected at baseline. Cortical cataract was defined as 4/16 or greater opacity; progression was defined as a more than 2/16 increase. Nuclear cataract was defined as Wilmer standard grade 2 or higher; progression was defined as more than 0.5 increase. Posterior subcapsular (PSC) cataract was defined as opacity 1 mm2 or greater; progression was defined as greater than 1 mm2 increase. ● RESULTS: Of the 3,040 participants eligible to attend follow-up examinations, 2,594 (85% of those eligible) participated. The mean age of participants at follow-up was 62.5 years, and 55% were female. The percentage of patients who had at least one lens extracted over 5 years increased from 0.5% of those aged 40 to 49 years at baseline to 35.7% of those aged 80 years or more at baseline. The overall incidence of the three types of cataract was as follows: cortical 7.7% (95% confidence limits [CL] ⴝ 5.8 –9.8), nuclear 16.4% (95% CL ⴝ 12.1–20.8), and PSC 7% (95% CL ⴝ 5.3– 8.7). The overall progression of cataract was cortical 14.3% (95% CL ⴝ 10.2–18.3), nuclear 19.3% (95% CL ⴝ 15.9 – 22.7), and PSC 20% (95% CL ⴝ 8.7–31.1). The
Accepted for publication Aug 23, 2002. From the Marshfield Medical Research Foundation (C.A.M.), Marshfield Clinic, Marshfield, Wisconsin; and Centre for Eye Research Australia (C.A.M., B.N.M., P.N.D., H.R.T.), University of Melbourne, Melbourne, Australia. This study was supported by the National Health and Medical Research Council, the Victorian Health Promotion Foundation, the Dorothy Edols Estate, the Ansell Ophthalmology Foundation, the Jack Brockhoff Foundation, the Ian Potter Foundation, the Eye Ear Nose and Throat Research Institute, the Felton Bequest, the Hugh D. Williamson Foundation, and the Appel Family Bequest. Doctor McCarty was the recipient of the Wagstaff Research Fellowship in Ophthalmology from the Royal Victorian Eye and Ear Hospital. Inquiries to Catherine McCarty, PhD, MPH, Marshfield Medical Research Foundation, Marshfield Clinic, 1000 N. Oak Ave (ML2), Marshfield, WI 54449; fax: (715) 389-3880; e-mail: mccartyc@mmrf. mfldclin.edu
10
©
2003 BY
incidence and progression rates increased significantly by age, but the rates were not significantly different by sex. ● CONCLUSIONS: These cataract incidence data confirm the public health importance of cataract in Australia. The data also support the need to plan both primary prevention program and adequate surgical services to meet the anticipated increase in demand with the aging population. (Am J Ophthalmol 2003; 136:10 –17. © 2003 by Elsevier Inc. All rights reserved.)
C
ATARACT SURGERY IS THE MOST COMMONLY PER-
formed ophthalmic procedure in Australia,1 and the amount of cataract surgery is likely to double in the next 20 years with the aging of the population. We have shown previously in the Visual Impairment Project, a cross-sectional, population-based study, that although there are a number of identifiable risk factors for the prevalence of cataract,2 no segment of the adult Melbourne population was observed to be systematically underserviced in terms of cataract surgery.3 From these same data we estimated that the projected need for cataract surgery varies by level of opacity, visual acuity, and patient concern about vision.4 Relatively few data are available regarding the incidence of cataract5–10 or cataract surgery.11 The incidence of cataract has been shown in these four studies to be highly age dependent, but the studies differ as to whether incidence rates are higher among male or female populations. Data from the Beaver Dam Eye Study indicate that the location of lens opacity in the posterior subcapsular (PSC) region, as opposed to the cortex or nucleus, is the most important predictor of subsequent cataract extraction.11 As acknowledged by all of the authors, incidence data are useful for planning future health service delivery needs, in conjunction with population projections, and also for planning future intervention studies. The purpose of this study was to quantify the 5-year incidence of cataract and cataract surgery in the Melbourne Visual Impairment Project cohort.
ELSEVIER INC. ALL
RIGHTS RESERVED.
0002-9394/03/$30.00 PII S0002-9394(02)01844-5
METHODS THE MELBOURNE VISUAL IMPAIRMENT PROJECT IS A POPU-
lation-based study of eye disease, the methods for which have been published previously.12 Recruitment and baseline examinations for this prospective cohort study were conducted from 1992 to 1994. Nine pairs of census collector districts from the Melbourne Statistical Division were randomly selected from which to recruit eligible residents via a household census. Eligible residents were defined as permanent residents aged 40 years and older. At the time of the household census, basic demographic details including age, sex, race, education level, country of birth, language spoken at home, and use of eye care services, were collected. Appointments were scheduled at locally established test sites. The protocol was approved by the Human Research and Ethics Committee of the Royal Victorian Eye and Ear Hospital and conforms to the Declaration of Helsinki for research involving human subjects. In 1997, a 5-year follow-up examination commenced.13 All participants were traced, and participants living in the area were invited to again attend locally established test sites for the follow-up examinations. The same examination protocol was used with redundant questions eliminated. Reported deaths were confirmed through the National Death Index at the Australian Institute for Health and Welfare in Canberra. The following information was collected for each participant: presenting and best-corrected visual acuity on a 4 m logarithm of the minimal angle of resolution (LogMAR) chart, LogMAR reading vision at a distance comfortable for the participant, Humphrey Fastpac 24-2 visual fields (Humphrey Instruments Inc., San Leandro, California, USA), Tonopen (Oculab, Calendale, California, USA) intraocular pressure, personal health history, clinical slit-lamp examination, and photography of the lens and fundus. The clinical examination and photography of the lens were performed after pupil dilation to a minimum of 6 mm with tropicamide. Cataract was graded clinically and on photographs according to the Wilmer protocol.14 The cortex was assessed on retroillumination and the nucleus was assessed with a slit lamp. Cortical cataract was defined as 4/16 or greater opacity. Nuclear cataract was defined as Wilmer standard grade 2.0 or higher. Posterior subcapsular cataract was defined as PSC opacity greater than or equal to 1 mm2. Photogrades were used when available, and the agreement between photogrades and clinical grades was shown to be very high for all three types of opacity.2 Participants were queried about the year that any cataract extraction had been performed. The treating ophthalmologist was contacted to provide information on the type of opacity present before cataract extraction. Prior cataract surgery was confirmed on clinical examination and the status of the capsule, if present, was noted. VOL. 136, NO. 1
CATARACT INCIDENCE
IN
Patients were classified as having incidence cataract or progression of lens opacity on the basis of that event occurring in either eye, regardless of the status of the fellow eye at baseline. Except where indicated in the results, the specific cataract types are not mutually exclusive, that is, a person may be classified as having any, none, or all three of the cataract types. For each of the separate cataract types, the denominator for the incidence estimates included those participants who did not have that specific cataract type (as defined in the previous paragraph) at baseline in either eye. Given these definitions, it is possible that someone was classified as having incident cataract, but not classified as having opacity that progressed because the progression was not great enough to fit into the definition for progression. For each specific type of cataract, the numerator for incidence estimates included participants who developed a specific cataract type over the time period as well as those who had cataract surgery (reportedly for that type of cataract according to the treating ophthalmologist) during the time period. Progression of lens opacity was considered separately to cataract. Thus, a patient did not necessarily have to have a “cataract” to be classified as having a lens opacity that had progressed. Patients were considered to be at risk of having their lens opacity progress if they had opacity greater than 0 at baseline, if they did not have cataract surgery in both eyes over the intervening time period and if they did not have the highest level of opacity at baseline so that there was no possibility of progression. Cataract surgery was not included in the definition of progression of lens opacity. To assess measurement error associated with photograding of the lens, a sample of photographs from baseline was regraded by one of the photograders (P.N.D.) at follow-up. Interview data were entered directly into a Paradox database, while clinical data were entered twice and verified. SAS (SAS Software, Cary, North Carolina, USA) was used for all statistical analyses, and P ⬍ .05 was considered statistically significant. Intraclass correlation coefficients were calculated to assess agreement between graders. Ninety-five percent confidence limits (CL) around the incidence estimates were calculated according to Cochran15 to account for the cluster sampling design.
RESULTS TWO-HUNDRED THIRTY-ONE (7%) OF THE PARTICIPANTS
examined at baseline had died before the 5-year follow-up examinations.14 Participants with presenting visual acuity less than 6/12 at baseline were 2.3 times as likely to have died in the intervening 5 years. Eighty-five percent (2,594/ 3,040) of the remaining eligible group participated in the incidence study. The mean age of the participants at follow-up was 62.5 years (SD ⫽ 10.9 years) and 1,421 VISUAL IMPAIRMENT STUDY
11
12
AMERICAN JOURNAL
OF
No cataract-no cataract No cataract-cataract No cataract-surgery Cataract-cataract Cataract-surgery Surgery-surgery Total
*Participants potentially at risk for incident cataract. Normal font ⫽ cortical, bold font ⫽ nuclear, italic font ⫽ posterior subcapsular. The shaded area indicates participants potentially at risk for incident cataract.
2,786, 2,871, 3,033 158, 68, 59 32, 27, 47 179, 184, 31 25, 30, 10 51, 51, 51 3,231, 3,231, 3,231 565, 593, 620 45, 17, 13 7, 5, 9 35, 28, 6 5, 7, 3 0, 0, 0 657, 650, 651 148, 158, 196 26, 8, 6 6, 6, 11 30, 38, 2 7, 7, 2 12, 12, 12 229, 229, 229 23, 38, 43* 11, 4, 7* 10, 5, 14* 24, 16, 8 7, 12, 3 39, 39, 39 114, 114, 114 1, 9, 9* 4, 2, 3* 1, 4, 3* 11, 16, 4 6, 4, 2 0, 0, 0 23, 35, 21 67, 227, 51* 26, 29, 5* 0, 0, 0* 79, 86, 11 0, 0, 0 0, 0, 0 172, 342, 67 63, 77, 83* 42, 6, 22* 0, 0, 0* 0, 0, 0 0, 0, 0 0, 0, 0 105, 83, 105 1,897, 1,756, 2,009* 0, 0, 0* 0, 0, 0* 0, 0, 0 0, 0, 0 0, 0, 0 1,897, 1,756, 2,009
Baseline
22, 13, 22* 4, 2, 3* 8, 7, 10* 0, 0, 0 0, 0, 0 0, 0, 0 34, 22, 35
SurgerySurgery CataractSurgery CataractCataract No CataractSurgery No CataractCataract
Follow-up
OPHTHALMOLOGY
No CataractNo Cataract
TABLE 1. Lens Status in Each Eye of Participants at Baseline and Follow-up
Died
Missing Data for Both Eyes
Total
(55%) were female. The mean time elapsed between baseline and follow-up examinations was 4.5 years (range 4 –7; SD ⫽ 0.64). Of the 2,594 participants examined at follow-up, 2,392 (92%) had lens data available for both eyes at follow-up. The status of lens data for the three types of opacities in all 3,231 participants with baseline lens data are summarized in Table 1. The shaded area indicates those who were at risk of incident cataract. However, the actual number of persons at risk for a particular type of cataract is slightly less owing to data missing for the only eye at risk of incident cataract. For example, if a person had no cortical cataract in one eye and previous cataract surgery in the other eye at baseline, and the first eye did not have a gradeable photograph at follow-up, this person would not be included in the denominator for incident cortical cataract at follow-up. It is not possible from the information presented in Table 1 to infer the number of people at risk for progression of lens opacity, because some participants with no cataract at baseline had some lens opacity and were at risk of progression. The intraclass correlation for the agreement between the graders of baseline and follow-up cortical opacities was 0.95, and the agreement between the two graders of the follow-up cortical opacities was 0.85. The intraclass correlation for the agreement of nuclear opacity grading between baseline and follow-up was 0.78, whereas the agreement between the two graders at follow-up was 0.93. The intraclass correlation for the agreement between baseline and follow-up grading of PSC opacities was 0.86. The progression of the three types of lens opacities was defined to be greater than the largest (that is, most conservative) standard error for either of the graders, which was greater than the standard error of the difference between the graders: a change of more than 2/16 for cortical, a change of more than 0.5 for nuclear, and a change of more than 1 mm2 for PSC. The overall rate of cataract extraction in at least one eye was 5.1% (95% CL ⫽ 3.9 – 6.2). The rate increased significantly by age (Figure 1), but not by sex (P ⬎ .05, data not presented). The distribution of incident cataract type varied by age somewhat; however, pure incident nuclear cataract was the most common type in all age categories, followed by mixed cataract (Figure 1). The rate of cataract extraction was fairly steady since the baseline examination (data not presented), suggesting that participation in the study did not have any significant impact on cataract extraction. The rate of cataract extraction increased nearly linearly with the level of opacity at baseline and was higher overall for PSC opacity than cortical or nuclear opacity (Figures 2– 4). Of the 2,294 participants without cataract in either eye at baseline who had data at follow-up, 614 (26.8%) had developed cataract at follow-up. As we found at baseline,2 there was relatively little overlap of cataract type given the definition of cataract that was employed. Pure nuclear
JULY 2003
FIGURE 1. Age-specific incidence of any cataract or cataract surgery in at least one eye. PSC ⴝ posterior subcapsular.
FIGURE 2. Cumulative 5-year incident cataract extraction by level of cortical opacity at baseline.
FIGURE 3. Cumulative 5-year incident cataract extraction by level of nuclear opacity at baseline.
cataract was the predominant type, followed by pure cortical cataract (Figure 5). The age- and sex-specific incidence and progression rates of mixed cortical (Table 2), nuclear (Table 3), and PSC cataract (Table 4) reveal significant increases by age. VOL. 136, NO. 1
CATARACT INCIDENCE
IN
FIGURE 4. Cumulative 5-year incident cataract extraction by level of posterior subcapsular (PSC) opacity at baseline.
FIGURE 5. The distribution and overlap of the three cataract types in the 614 incident cases of cataract. PSC ⴝ posterior subcapsular.
Although some of the incidence and progression rates were higher among women, these differences were not statistically significant. Overall, the incidence of nuclear cataract was twice that of cortical or PSC cataract and this difference was statistically significant. The overall progression rates of the three types of cataract were not significantly different. The progression rate was higher than the incidence rate. There were only 45 cases of PSC cataract at baseline with available follow-up data—too few to calculate age- and sex-specific rates. The overall progression rate in this group was 20.0% (95% CL ⫽ 8.7–31.3). We calculated the incidence of cortical and nuclear cataract using two alternative definitions for each cataract type. The incidence of cortical cataract 2/16 or greater was 18.3% (95% CL ⫽ 14.1–22.5), 3/16 or greater was 15.8% (95% CL ⫽ 12.9 –18.6), and 6/16 or greater was 5.3% (95% CL ⫽ 3.9 – 6.7). The incidence of nuclear cataract standard 1.5 or greater was 45.3% (95% CL ⫽ 41.4 – 49.2) and standard 2.5 or greater was 8.9% (95% CL ⫽ 5.9 – 12.0). We were unable to conduct sensitivity analyses for PSC owing to the relatively small number of cases. VISUAL IMPAIRMENT STUDY
13
TABLE 2. Age and Sex-specific Incidence and Progression of Cortical Cataract Men Age Group (years)
Incidence 40–49 50–59 60–69 70–79 80⫹ Total Progression 40–49 50–59 60–69 70–79 80⫹ Total
Number at Risk
Women Percent (95% CL)
259 322 265 109 15 970
1.9 (0.34–3.5) 4.0 (2.3–5.7) 8.3 (4.5–12.2) 11.9 (3.9–19.9) 13.3 (0–29.9) 5.7 (3.6–7.8)
8 28 52 55 9 152
12.5 (0.57–24.4) 14.3 (0–29.9) 9.6 (0.03–19.2) 12.7 (5.4–20.1) 11.1 (0–33.6) 11.8 (5.7–18.0)
All
Number at Risk
Percent (95% CL)
Number at Risk
Percent (95% CL)
343 401 278 100 21 1,143
3.5 (0.84–6.2) 9.0 (6.3–11.6) 16.2 (13.2–19.2) 12.0 (5.1–18.9) 9.5 (0–22.2) 9.4 (7.5–11.2)
602 723 543 209 36 2,313
2.8 (1.1–4.5) 6.8 (4.7–8.8) 12.3 (10.5–14.2) 12.0 (6.7–17.3) 11.1 (2.5–19.7) 7.7 (5.8–9.5)
12 38 89 54 27 220
8.3 (0–24.0) 23.7 (13.9–33.5) 14.6 (7.3–21.9) 18.5 (5.4–31.7) 7.4 (0–16.3) 15.9 (11.0–20.8)
20 66 141 109 36 372
10.0 (0.88–19.1) 19.7 (11.6–27.8) 12.8 (6.5–10.1) 15.6 (8.4–22.8) 8.3 (0.18–16.5) 14.3 (10.2–18.3)
CL ⫽ confidence limits.
TABLE 3. Age and Sex-specific Incidence and Progression of Nuclear Cataract
Age Group (years)
Incidence 40–49 50–59 60–69 70–79 80⫹ Total Progression 40–49 50–59 60–69 70–79 80⫹ Total
Men Number at Risk
Women
All
Percent (95% CL)
Number at Risk
Percent (95% CL)
Number at Risk
Percent (95% CL)
259 330 273 114 9 985
1.5 (0.44–2.7) 7.0 (3.1–10.8) 26.0 (18.7–33.3) 55.3 (47.9–62.7) 66.7 (36.4–96.9) 17.0 (13.0–20.9)
343 405 284 89 13 1,134
2.0 (0.13–4.0) 7.7 (3.0–12.3) 28.9 (22.6–35.2) 57.3 (46.2–68.4) 76.9 (49.3–10.5) 15.9 (10.6–21.4)
602 735 557 203 22 2,119
1.8 (0.48–3.2) 7.4 (3.4–11.3) 27.5 (21.6–33.4) 56.2 (50.3–62.1) 72.7 (51.8–93.6) 16.4 (12.1–20.8)
259 330 283 134 18 1,024
7.7 (4.4–11.1) 12.7 (8.8–16.6) 20.9 (16.6–25.1) 47.0 (37.7–56.3) 50.0 (20.0–80.0) 18.9 (16.0–21.7)
345 405 303 123 33 1,209
10.4 (7.4–13.4) 12.6 (8.6–16.6) 26.1 (19.5–32.6) 47.9 (39.3–56.7) 39.4 (23.5–55.3) 19.7 (15.3–24.1)
604 735 586 257 51 2,233
9.3 (6.6–12.0) 12.7 (9.1–16.3) 23.6 (18.7–28.4) 47.5 (39.9–55.0) 43.1 (35.4–50.8) 19.3 (15.9–22.7)
CL ⫽ confidence limits.
levels of ultraviolet (UV-B) radiation in these environments. We have shown previously that higher levels of UV-B are significantly related to the prevalence of cortical cataract.2 Unexplainably, all of the observed cortical cataract incidence rates are substantially less than the 28.2% incidence rate that was observed among older participants aged 65 to 74 in the Italian-American Cataract Study.5 The cumulative nuclear cataract standard 2.0 or greater incidence rate of 16.4% observed in the current study is higher than in previous studies (Table 5). These differences may be explained by the different definitions of
DISCUSSION THESE DATA CONFIRM THAT CATARACT IS A MAJOR PUB-
lic health problem in Australia that will only increase as the population ages. The cumulative incidence rate of cortical cataract of 7.7% when defined as 4/16 or greater that was observed in the current study is similar to that of the Beaver Dam Eye Study,8 and the Longitudinal Study of Cataract7 (Table 5). When considering the cutoff of 3/16 or more, the incidence rate in the Visual Impairment Project is closer to that of the white participants in the Barbados Eye Study.10 That could be due to the higher 14
AMERICAN JOURNAL
OF
OPHTHALMOLOGY
JULY 2003
TABLE 4. Age and Sex-specific Incidence of Posterior Subcapsular Cataract
Age Group (years)
Men Number at Risk
40–49 50–59 60–69 70–79 80⫹ Total
258 329 274 127 16 1,004
Women Percent (95% CL)
All
Number at Risk
Percent (95% CL)
Number at Risk
Percent (95% CL)
344 404 297 118 31 1,194
1.2 (0.41–1.9) 4.2 (2.4–6.0) 11.1 (7.5–14.7) 21.2 (15.7–26.7) 16.1 (7.0–25.3) 7.0 (5.7–8.4)
602 733 571 245 47 2,198
2.3 (1.4–3.3) 4.1 (2.4–5.8) 8.1 (5.7–10.4) 22.9 (16.5–29.2) 17.0 (7.5–26.5) 7.0 (5.3–8.7)
3.9 (1.8–6.0) 4.0 (0.39–7.5) 4.7 (2.4–7.1) 24.4 (14.5–34.3) 18.8 (0.34–37.2) 7.0 (4.4–9.5)
CL ⫽ confidence limits.
TABLE 5. Published Incidence Rates of Cataract Study
Study Population
Time Frame
Cataract Definition
Melbourne Visual Impairment Project
Cluster, random sample of Melbourne residents aged 40⫹ at baseline, 1992– 1994
5 years
Italian-American Cataract Study Group5 Longitudinal Study of Cataract6,7
Follow-up of clinic-based case-control study in Parma, Italy, patients aged 45–79 at baseline, 1987–1989 Follow-up of clinic-based Lens Opacities Case-Control Study, patients aged 40–70⫹ at baseline, 1989–1993 Rural Wisconsin residents aged 43–84 at baseline, 1988–1990
5 years
Cortical ⱖ2/16 Cortical ⱖ3/16 Cortical ⱖ4/16 Cortical ⱖ6/16 Nuclear ⱖstandard 1.5 Nuclear ⱖstandard 2.0 Nuclear ⱖstandard 2.5 PSC ⱖ1 mm2 Any cataract Any and VA ⱕ02 LogMAR Cortical, any, aged 65–74 years Nuclear, any, aged 65–74 years PSC, any, aged 65–74 years Cortical, LOCS III ⱖ2.0 Nuclear PSC Cortical, ⱖ5% lens surface Nuclear, ⱖgrades 4 and 5 PSC, ⱖ5% grid segment Any lens opacity and VA ⱕ0.2 LogMAR
18.3% (14.1–22.5) 15.8% (12.9–18.6) 7.7% (5.8–9.5) 5.3% (3.9–6.7) 45.3% (41.4–49.2) 16.4% (12.1–20.8) 8.9% (5.9–12.0) 7.0% (5.3–8.7) 28.1% (23.5–32.7) 10.6% (8.3–13.0) 28.2% (SE ⫽ 3.2) 11.5% (SE ⫽ 1.5) 9.6% (SE ⫽ 1.3) 7.7% (SE ⫽ 1.2) 7.7% (SE ⫽ 1.5) 4.3% (SE ⫽ 0.9) 8.0% (not available) 13.1% (not available) 3.4% (not available) 9.0% (6.7–11.3)
Cortical, LOCS II scores ⱖ2 Nuclear, LOCS II scores ⱖ2 PSC, LOCS II scores ⱖ2
Black:22.2, White:16.2 Black:9.2, White:8.7 Black:3.3, White:3.9
Beaver Dam Eye Study8 Italian Study9
Barbados Eye Study10
5 years
5 years
Priverno participants in cardiovascular study, aged 45–69 years at baseline, 1987 Population-based sample of Barbados residents aged 40–84 at baseline, 1987–1992.
7 years
4 years
Cataract Rate (95% CL)
LOCS ⫽ Lens Opacity Case-Control Study; LogMAR ⫽ logarithm of the minimum angle of resolution; PSC ⫽ posterior subcapsular; VA ⫽ visual acuity.
nuclear cataract in the various studies, because when a more conservative definition was employed with our data, the incidence rate in the current study is similar to that in other studies. The cumulative PSC incidence rate of 7.0% in the current study is slightly higher than has been observed in VOL. 136, NO. 1
CATARACT INCIDENCE
IN
previous studies (Table 5). This difference may again reflect grading and classification differences. The data in Table 5, including the sensitivity analyses for the current study, reveal the large effect on incidence of different cut-off points for cataract classification. These data highlight the need for an objective cataract grading scheme and VISUAL IMPAIRMENT STUDY
15
definition that can be employed by all researchers to allow more valid comparisons between studies. The 28% 5-year incidence of any cataract in our population is significantly higher than the cumulative 7-year incidence of 9% reported in a sample of Italians aged 45 to 69 years.9 However, the definition of cataract employed by these researchers included a criterion of decreased visual acuity associated with age-related lens opacity. When we included the additional criterion of visual acuity less than 0.2 LogMAR in our definition of cataract, we found the incidence of any cataract to be 10.6% (95% CL ⫽ 8.3–13.0), which is not significantly different from the finding in the Italian study.9 Women have been reported to have significantly higher incidence rates of nuclear cataract in the Beaver Dam Eye Study8 and the Barbados Eye Study,10 whereas the Longitudinal Study of Cataract6 failed to find any significant difference in incidence of nuclear cataract between sexes. Although the incidence of nuclear cataract was slightly higher among women than men in our study, this finding was not statistically significant. Prior research has suggested that the increased prevalence of nuclear cataract that has been observed among women may be due to hormonal differences and that hormone replacement therapy may be protective against nuclear cataract;17 however, our baseline data did not support this finding2 and no prospective data are available. The threefold difference in cataract incidence that was observed in the Visual Impairment Project with varying definitions of cataract, as well as the difference in rates among studies with similar populations, highlights the lack of objectivity in current definitions of cataract. This problem is further compounded by the use of many different cataract grading schemes. One solution to this problem could be the inclusion of a visual acuity or visual function criterion along with lens opacity to define cataract. Another possible solution would be the use of objective systems, such as digital measurements, to classify lens opacity. Clearly, standardization of measurement technique and objectivity in defining cataract are necessary to allow comparisons of results between studies. The strengths of this study include the sampling strategy and relatively high response rate, which resulted in a study sample that is representative of the Melbourne population aged 40 years and older. A weakness of any incidence study with no data at intervening time points is that the estimation of incidence and progression could be affected by differential mortality. That would lead to a slight underestimate of the incidence and progression rates if participants who died were more likely to develop cataract or to have their opacity progress before their death. We showed previously that participants with cortical cataract at baseline were slightly more likely to die in the intervening 5 years, but this finding was not significant in multivariate analyses.14 An advantage of incidence data for cataract surgery over cross-sectional data is that the 16
AMERICAN JOURNAL
information can better be used to plan for eye care services because it is easier to estimate annual demand. Another limitation of these data is that information about the status of lenses immediately before cataract extraction was not available. Estimates of cataract surgery rates could also be influenced by participation in the baseline examination. Specifically, patients may be more likely to seek cataract extraction if they had been told during the ophthalmic examination that they had significant lens opacity. However, we found no obvious impact of participation in baseline examinations because the cataract extraction rates were fairly constant over the time between baseline and follow-up examinations. That is not unexpected, as we have shown previously in this cohort that no one under the age of 70 years had vision less than 6/12 due to cataract, and the rate of vision impairment due to cataract rose to 11% of patients aged 90 years and older.16 In conclusion, these cataract incidence data confirm the public health importance of cataract in Australia. The data also support the need to plan both primary prevention programs to encourage reduction in known environmental risk factors such as cigarette smoking and ocular UV-B radiation exposure and adequate surgical services to meet the anticipated increase in demand with the ageing population. ACKNOWLEDGMENTS
The authors acknowledge the assistance of Sharon Bayley, Marie Bissinella, Shayne Brown, Dr. Van Lansingh, Dr. Patricia Livingston, Claire McKean, Dr. Yury Stanislavsky, Dr. Mylan Van Newkirk, Cathy Walker, and Matthew Wensor in the data collection.
REFERENCES 1. Keeffe JE, Taylor HR. Cataract surgery in Australia 1985– 1994. Aust N Z J Ophthalmol 1996;24:313–317. 2. McCarty CA, Mukesh BN, Fu CL, Taylor HR. The epidemiology of cataract in Australia. Am J Ophthalmol 1999; 128:446 –465. 3. McCarty CA, Keeffe JE, Taylor HR. The need for cataract surgery: projections based on lens opacity, visual acuity, and personal concern. Br J Ophthalmol 1999;83:62–65. 4. McCarty CA, Nanjan MB, Taylor HR. Operated and unoperated cataract in Australia. Clin Exp Ophthalmol 2000;28: 77–82. 5. Italian-American Cataract Study Group. Incidence and progression of cortical, nuclear, and posterior subcapsular cataracts. Am J Ophthalmol 1994;118:623–631. 6. Leske MC, Chylack LT, Wu S-Y, et al. Incidence and progression of nuclear opacities in the Longitudinal Study of Cataract. Ophthalmology 1996;103:705–712. 7. Leske MC, Chylack LT, He Q, et al. Incidence and progression of cortical and posterior subcapsular opacities. The OF
OPHTHALMOLOGY
JULY 2003
8.
9.
10.
11.
12.
Longitudinal Study of Cataract. Ophthalmology 1997;104: 1987–1993. Klein BEK, Klein R, Lee KE. Incidence of age-related cataract. The Beaver Dam Eye Study. Arch Ophthalmol 1998;116:219 –225. Cedrone C, Culasso F, Cesareo M, et al. Prevalence and incidence of age-related cataract in a population sample from Priverno, Italy. Ophthalmic Epidemiol 1999;6:95–103. Leske MC, Wu S-Y, Nemesure B, Li X, Hennis A, Connell AMS, and the Barbados Eye Studies Group. Incidence and progression of lens opacities in the Barbados Eye Studies. Ophthalmology 2000;107:1267–1273. Klein BEK, Klein R, Moss SE. Incident cataract surgery. The Beaver Dam Eye Study. Ophthalmology 1997;104:573– 580. Livingston PM, Carson CA, Stanislavsky YL, Lee SE, Guest
VOL. 136, NO. 1
CATARACT INCIDENCE
IN
13.
14. 15. 16.
17.
CS, Taylor HR. Methods for a population-based study of eye disease: the Melbourne Visual Impairment Project. Ophthalmic Epidemiol 1994;1:139 –148. McCarty CA, Nanjan MB, Taylor HR. Vision impairment predicts five year mortality. Br J Ophthalmol 2001;85:322– 326. Taylor HR, West SK. A simple system for the clinical grading of lens opacities. Lens Res 1988;5:139 –148. Cochran WG. Sampling techniques. New York: John Wiley & Sons, 1977, 249 –271. Weih LM, VanNewkirk MR, McCarty CA, Taylor HR. Age-specific causes of bilateral visual impairment. Arch Ophthalmol 2000;118:264 –269. Klein BEK, Klein R, Ritter L. Is there evidence of an estrogen effect on age-related lens opacities? The Beaver Dam Eye Study. Arch Ophthalmol 1994;112:85–91.
VISUAL IMPAIRMENT STUDY
17